Introduction Faecal calprotectin (FC) has become a useful marker of mucosal healing, with studies showing raised levels predictive of relapse^1,2,3. These studies did not assess mucosal healing however, so did not compare calprotectin with mucosal appearance as predictors of relapse.

Methods In the CODA (Colitis Once Daily Asacol®) trial of once daily (OD) Asacol® (three 800 mg tablets) vs one 800mg tablet taken three times daily (TDS), 213 UC patients in remission for >4 weeks, but relapse in the past 2 years, were recruited. Baseline FC (Phical ELISA kit) was collected and rectal sigmoidoscopy (sig) score at baseline, and relapse or 1 year (using the modified Baron score: 0 = normal; 1 = erythema, decreased vascular pattern; 2 = marked erythema, absent vascular pattern, friability, erosions; 3 = spontaneous bleeding, ulceration). At entry patients had no symptoms of active disease, with a sig score of 0 or 1. Follow-up was for 1 year or until relapse (symptoms of active disease with a sig score of 2 or 3). Demographic factors, concomitant drugs, FC, sig score, CRP, and adherence were evaluated in a Cox regression model of time to relapse.

Results (Shown as median[IQR] unless stated otherwise). Remission duration prior to entry was 6.0 [3–12] months. Disease extent was extensive (30.0%), Lt. sided (54.9%), proctitis (13.6%). At entry FC was 78mg/kg stool [23.3–159.4], sigmoidoscopy score 0 (70.9%), 1 (29.1%). All were taking mesalazine, and 11.7% thiopurines. Baseline FC was higher if sig score was 1 (109[38–335]) than if 0 (62[21–120], p = 0.001, but did not differ according to disease extent or medication (including aspirin (n = 18) and occasional NSAIDs (n = 6). Remission rates at one year were 62% overall (68.9% in OD and 55.5% in TDS group). Factors associated with time to relapse were explored in a Cox proportional hazards model, with baseline FC dichotomised at 150mg/kg stool (as by Costa et al. Relapse risk was 2.5 times higher in those with baseline sig score 1 compared to score 0 (95% CI 1.32–4.76, p = 0.005). Age, concomitant medication, previous duration of remission, disease extent, smoking, trial medication adherence and baseline FC (cut off at 150 mg/kg stool or as a continuous variable) did not remain in the final model.

Conclusion In this study, sigmoidoscopy appearance at baseline was the sole factor predicting relapse over 1 year of maintenance mesalazine 2.4g, whereas calprotectin level was not a predictor, perhaps because of wide variability in this group of patients. FC may have more value in measuring relative change in individual patients.

Disclosure of Interest A. Hawthorne Grant/Research Support from: CODA study supported by unrestricted grant from Warner Chilcott, D. Gillespie: None Declared, C. Probert: None Declared

References

1. Gisbert IBD 2009; 15:1190.

2. Tibble Gastroenterol 2000; 119:15.

3. Costa Gut 2005; 54:364.